Early warning model construction method and system for urban safety risk assessment benchmark

Abstract

The invention discloses an early warning model construction method and system for urban safety risk assessment benchmark, and the method comprises the steps: updating urban safety historical data, determining a risk threshold value, determining risk data according to the risk threshold value, obtaining a historical data matrix according to an urban database, and carrying out the early warning of the urban safety risk assessment benchmark. Determining the assessment reference value according to the historical matrix and the risk data, determining a risk assessment value according to the risk data and the risk threshold value, and optimizing the urban safety risk assessment reference early warning model according to the risk assessment value deviation. And inputting to-be-assessed city safety data into the optimized city safety risk assessment reference early warning model for risk assessment and risk early warning. The method not only can improve the efficiency and accuracy of urban safety risk assessment benchmark management, but also has good interpretability, and can be directly applied to an early warning model system of the urban safety risk assessment benchmark.

Description

Technical Field

[0001] The present invention relates to the technical field of urban safety management, and particularly to a method and system for constructing an early warning model for urban safety risk assessment benchmarks. Background Art

[0002] With the acceleration of the urbanization process, urban safety risks have become increasingly complex and diverse. At the same time, the occurrence of these risks often has characteristics such as suddenness, complexity, and chain reaction. Therefore, urban safety risk assessment has become a key link in urban management.

[0003] Traditional urban safety risk assessment methods often rely on manual experience and qualitative analysis, lacking scientificity and accuracy. At the same time, urban safety data is massive and complex. Most existing urban safety risk assessment systems adopt static risk assessment models, which cannot reflect the dynamic changes of urban safety risks in real time. In addition, these systems and assessment methods are mostly based on single indicators or simple weighted calculations, often ignoring the correlation between risks, resulting in incomplete and in-depth risk assessment results. At the same time, existing systems mostly focus on post-event analysis and lack the ability to monitor and early warn potential risks in real time. Therefore, the present invention proposes a method and system for constructing an early warning model for urban safety risk assessment benchmarks, which can achieve a comprehensive perception, accurate assessment, and real-time early warning of urban safety risks by constructing a multi-level and dynamic assessment model, overcome the deficiencies of existing early warning models, and realize the comprehensive management and dynamic optimization of urban safety risks. Summary of the Invention

[0004] The object of the present invention is to provide a method and system for constructing an early warning model for urban safety risk assessment benchmarks.

[0005] To achieve the above object, the present invention is implemented according to the following technical solutions: The present invention includes the following steps: Construct an urban database: update historical urban safety data and determine risk thresholds; Construct a risk identification layer: determine risk data according to the risk thresholds; Determine the evaluation benchmark value: obtain a historical data matrix according to the urban database, and determine the evaluation benchmark value according to the historical matrix and the risk data; Construct a risk assessment layer: determine risk valuations according to the risk data and the risk thresholds; Optimize the model: optimize the urban safety risk assessment benchmark early warning model according to the risk valuation deviation; Construct a risk early warning layer: conduct risk early warning according to the risk valuations and the evaluation benchmark value.

[0006] Furthermore, the method for constructing the urban database includes: Obtain the historical data of urban safety, classify the data using the clustering method, store the historical data of urban safety in different partitions according to the classification results, and store the historical data of urban safety in the storage partitions in segments according to the collection time; Set up a preliminary screening window for urban safety data, determine the parameters of the preliminary screening window, and slide and update the historical data of urban safety in the urban database in the preliminary screening window of urban safety data according to the time series; the parameters of the preliminary screening window include the update frequency and the window length; Determine the update weights of various types of data, and divide various types of data into type I data and type II data according to the data update weights; Obtain the newly generated historical data of urban safety, and determine the preliminary screening result of the data according to the data deviation between the newly generated historical data of urban safety and the mean value of the historical data of urban safety in the preliminary screening window of urban safety data; the preliminary screening result of the data includes further determining the data update situation and not updating the historical data; The steps of further determining the data update situation include: Calculate the category similarity between the newly generated historical data of urban safety in different partitions and the historical data of urban safety in this category, determine the historical data of urban safety to be replaced in this partition according to the category similarity threshold, and extract the corresponding historical data group of urban safety according to the time stamp; Calculate the overall similarity between the newly generated historical data of urban safety and the historical data of urban safety in all categories, determine the update group of the historical data of urban safety according to the overall similarity threshold. When the number of update groups is less than 3, store the newly generated historical data of urban safety in the urban database according to the rules and do not replace the historical data of urban safety. Otherwise, add the time series weight to calculate the time series overall similarity, and replace the historical data of urban safety with the lowest time series overall similarity with the newly generated historical data of urban safety; Based on the updated urban database, use statistical methods to determine the risk threshold of urban safety data.

[0007] Furthermore, the method for calculating the similarity includes: Distinguish numerical data and semantic data, and set up a semantic evaluation set , is the evaluation set element corresponding to the semantic evaluation of the semantic data, is the evaluation degree, and the semantic numerical data is obtained through numerical conversion by the semantic function . Determine the semantic feature vector according to the semantic numerical data, and calculate the semantic similarity between the newly generated historical data of urban safety and the corresponding semantic feature vector of the historical data of urban safety. The expression is: , where is the semantic feature vector of the newly generated historical data of urban safety and the first The semantic feature vector of the historical data of urban security and the semantic similarity is the Manhattan distance between and is the maximum Manhattan distance between and all is the word frequency length of the corresponding semantic data is the word frequency length of the corresponding semantic data is the maximum length of the word frequencies of the corresponding semantic data in all ; Calculate the numerical similarity between the newly generated historical data of urban security and the numerical feature vector of the historical data of urban security. The expression is: , where is the numerical feature vector of the newly generated historical data of urban security and the th numerical feature vector of the historical data of urban security and the numerical similarity , are respectively the th numerical element value of the numerical feature vector of the newly generated historical data of urban security and the mean value of all numerical elements, , are respectively the th numerical element value of the numerical feature vector of the historical data of urban security and the mean value of all numerical elements, is the dimension of the feature vector; Thus, calculate the similarity between the newly generated historical data of urban security and the th group of historical data of urban security as , is the similarity weight.

[0008] Furthermore, the method for determining the evaluation benchmark value includes: Obtain the urban security data to be evaluated, classify the urban security data to be evaluated using a decision tree, screen the classification results according to the risk thresholds of different types of urban security data to obtain risk data, and extract the historical deviations of different types of data to obtain a historical deviation vector; the historical deviation is determined by the historical maximum value, historical minimum value, and historical mean value; Construct a historical data matrix based on the urban database; the column data of the historical data matrix is used to represent different statistical forms of the same type of data; the row data of the historical data matrix is used to represent different types of data in the same statistical form; Sort the risk data by category to obtain a risk vector, compare the risk vector with the historical data matrix to obtain a historical difference matrix, determine a risk threshold vector from the risk thresholds of different types of urban safety data, divide each element in each row of the historical difference matrix by the corresponding element in the risk threshold vector to obtain a weighted difference coefficient matrix, and determine the evaluation benchmark value for each category based on the weighted difference coefficient matrix and the historical deviation vector.

[0009] Furthermore, the method for constructing the risk assessment layer includes: Calculate the mean of the data correlation between risk data of the same category and risk data of other categories to obtain a risk weight; Calculate the deviation value of different category risk data from the corresponding risk threshold to obtain a risk deviation; Calculate the prior probability of urban safety data based on the historical data matrix and associated events, and use Bayes' theorem to calculate the corresponding risk probability based on different category risk data and the prior probability. Calculate the prior probability of urban safety data based on the historical data matrix and associated events, and use Bayes' theorem to calculate the corresponding risk probability based on the risk data and the prior probability Determine the risk valuation of different types of urban safety data based on the risk weight, risk deviation, and risk probability. The expression of the risk valuation function is: , where is the risk valuation of the th type of urban safety data, is the bias term of the risk valuation of the th type of urban safety data, used to adjust the valuation benchmark, is the risk weight of the th type of risk data, is the risk deviation of the th data index in the th type of risk data from the corresponding risk threshold, is the number of data indices of the th type of risk data, is the risk probability of the th type of risk data, , is the mean and standard deviation of the prior probability of the th type of urban safety historical data, is the regularization coefficient, is a non-zero constant.

[0010] Further, the method for optimizing the model includes: Determine the objective function for optimizing the early warning model according to the risk valuation deviation, and the expression is: , where is the objective function for optimizing the early warning model, is the evaluation benchmark value of the th type of urban safety data, is the number of categories of urban safety data, is the th number of indicators of the urban safety data, is the regularization parameter, is the hyperparameter of the weight of the interpretability term, is the th parameter of the model, is the number of model parameters; Update the existing parameter information by using Gaussian process regression, and the expression is: , where is the mean value of the function at , is the input point corresponding to the risk valuation function, is the non-linear mapping of the input data in the high-dimensional feature space, is the covariance matrix of the training data in the mapped feature space, which is calculated by the adaptive kernel function, is the non-linear mapping of the training data in the high-dimensional feature space, is the variance of the data Gaussian noise, is the identity matrix, is the target vector of the training data, is the covariance of the function at , is the adaptive kernel function, is the kernel function parameter for adaptive adjustment; Select the most promising set of hyperparameters for the next evaluation according to the acquisition function, and the acquisition function expression is: , where is the hyperparameter selection benchmark, is the objective function threshold, is the proposed hyperparameter, is the actual value of the objective function when using the hyperparameter , is the actual value of the objective function is the probability of the surrogate model at that time, is the objective function value Less than the threshold When the hyperparameter The probability of occurrence, Is the objective function value Less than the threshold The probability of, Is the objective function value Greater than the threshold When the hyperparameter The probability of occurrence; Update the parameters and evaluate the hyperparameters, repeat the above operations until the objective function value of the early warning model optimization is minimized, stop updating the hyperparameters, and output the urban safety risk assessment benchmark early warning model; Input the urban safety data to be evaluated into the optimized urban safety risk assessment benchmark early warning model to obtain the risk valuations and corresponding assessment benchmark values of different categories of urban safety data, conduct risk early warning according to the risk valuation deviation, and use the corresponding risk data, risk thresholds, assessment benchmark values, and risk valuations as the historical assessment set, and store the historical assessment set in partitions.

[0011] In a second aspect, an early warning model system for an urban safety risk assessment benchmark includes: Urban database module: used to store urban safety historical data, used to calculate the similarity of the urban safety historical data to update the urban safety historical data, used to determine the historical data matrix according to the urban safety historical data, used to determine the risk threshold, and used to store the historical assessment set; Risk identification module: used to obtain urban safety data and screen the urban safety data according to the risk threshold to determine risk data; Assessment benchmark value module: used to obtain the historical data matrix according to the urban database, and determine the assessment benchmark value according to the historical matrix and the risk data; Risk assessment module: used to calculate risk weights, risk deviations, and risk probabilities, and determine the risk valuations of different categories of urban safety data according to the risk weights, the risk deviations, and the risk probabilities Model optimization module: used to optimize the urban safety risk assessment benchmark early warning model according to the risk valuation deviation; Risk management module: used to view and manage the urban safety historical data and the historical assessment set, and conduct risk early warning according to the risk valuation and the assessment benchmark value.

[0012] The beneficial effects of the present invention are: The present invention is a method and system for constructing an early warning model for an urban safety risk assessment benchmark. Compared with the prior art, the present invention has the following technical effects: The present invention constructs a unified urban database to integrate multi-source data, solves the problems of data dispersion and lagging updates, and improves the data integration ability of the early warning model; the present invention dynamically determines the evaluation benchmark value based on the historical data matrix and risk data, improving the scientificity and accuracy of the evaluation; the present invention ensures the consistency between the evaluation result and the actual risk through risk valuation deviation feedback and dynamic optimization of model parameters; in addition, a risk early warning layer is constructed to achieve real-time monitoring and early warning of potential risks, which can greatly save resources and improve work efficiency, can realize the management of the urban safety risk assessment benchmark, can comprehensively, objectively and accurately obtain the safety risk status of urban areas, provide precise scientific guidance for urban safety risk control, and can meet the terminal management requirements of an early warning system for different urban safety risk assessment benchmarks and different users, having a certain universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a flowchart of the steps of a method for constructing an early warning model for an urban safety risk assessment benchmark of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The present invention will be further described below through specific embodiments. The illustrative embodiments and explanations of the present invention are used to explain the present invention, but do not limit the present invention.

[0015] A method and system for constructing an early warning model for an urban safety risk assessment benchmark of the present invention include the following steps: As Figure 1 shown, in this embodiment, it includes the following steps: Construct an urban database: update urban safety historical data and determine risk thresholds; Construct a risk identification layer: determine risk data according to the risk thresholds; Determine the evaluation benchmark value: obtain a historical data matrix according to the urban database, and determine the evaluation benchmark value according to the historical matrix and the risk data; Construct a risk assessment layer: determine a risk valuation according to the risk data and the risk thresholds; Optimize the model: optimize the early warning model for urban safety risk assessment according to the risk valuation deviation; Construct a risk early warning layer: conduct risk early warning according to the risk valuation and the evaluation benchmark value.

[0016] In this embodiment, the method for constructing the urban database includes: Obtain urban safety historical data, use the cluster clustering method for data classification, store the urban safety historical data in partitions according to the classification results, and store the urban safety historical data in the stored partitions in segments according to the collection time; Set up an initial screening window for urban safety data, determine the parameters of the initial screening window, and slide and update the urban safety historical data in the urban database according to the time series in the urban safety data initial screening window; the parameters of the initial screening window include the update frequency and the window length; Determine the update weights of various types of data, and divide various types of data into type I data and type II data according to the data update weights; Taking the urban database of a certain city as an example, the classification results and corresponding weights are 0.3 for natural disasters, 0.25 for accident disasters, 0.2 for public health, 0.15 for social security, and 0.1 for infrastructure. Among them, the urban safety historical data of natural disasters and accident disasters are classified as type I data, and the public health, social security, and infrastructure are classified as type II data; Obtain the newly generated urban safety historical data, calculate the first-class deviation of the newly generated urban safety historical data from the mean value of the urban safety historical data in the urban safety data initial screening window corresponding to type I data. When the first-class deviation is greater than the first-class deviation threshold, further determine the data update situation. Otherwise, calculate the second-class deviation of the newly generated urban safety historical data from the mean value of the urban safety historical data in the urban safety data initial screening window corresponding to type II data. When the second-class deviation is greater than the second-class deviation threshold, further determine the data update situation. Otherwise, do not update the historical data; Taking the newly generated urban safety historical data of a certain city as an example: Type I data: 1. Natural disasters (level 3 typhoon, typhoon affecting 60% of the urban area, daily rainfall of 250 mm, riverbed height increased by 6 m, number of waterlogging areas 15, waterlogging depth 0.3 m), 2. Accident disasters (1 chemical leakage incident in the industrial area); Type II data: 1. Public health (no special events), 2. Infrastructure (5 collapses of old houses, 10 malfunctions of urban elevators, 20 malfunctions of power operations, 2 gas leakage explosions, 1 bridge operation malfunction, 5 failures of water supply and drainage pipelines, normal traffic congestion index, normal number of emergency service calls, 50% increase in mobile network traffic, 1 damage to public facilities), 3. Public security (10 public transportation malfunctions, 15 theft incidents); The first-class deviation of the newly generated urban safety historical data from the mean value of the urban safety historical data in the urban safety data initial screening window corresponding to type I data is greater than the first-class deviation threshold, so the data update situation is further determined; The steps for further determining the data update situation include: Calculate the category similarity of the newly generated urban safety historical data and the urban safety historical data in different partitions in this category, determine the urban safety historical data to be replaced in this partition according to the category similarity threshold, and extract the corresponding urban safety historical data group according to the time stamp; Calculate the overall similarity of the newly generated urban safety historical data and the urban safety historical data in all categories, determine the urban safety historical data update group according to the overall similarity threshold. When the number of update groups is less than 3, store the newly generated urban safety historical data in the urban database according to the rules without replacing the urban safety historical data. Otherwise, calculate the temporal overall similarity by adding temporal weights, and replace the urban safety historical data with the lowest temporal overall similarity with the newly generated urban safety historical data; Based on the updated urban database, use statistical methods to determine the risk threshold of urban safety data.

[0017] In this embodiment, the method for calculating the similarity includes: Distinguish numerical data and semantic data, and set a semantic evaluation set , is the evaluation set element corresponding to the semantic evaluation of the semantic data, is the evaluation degree, and the semantic numerical data is obtained through numerical conversion by the semantic function . The expression is: , where is the semantic function, is the semantic coefficient, and the value range is determined by fitting; Determine the semantic feature vector according to the semantic numerical data, and calculate the semantic similarity between the newly generated urban safety historical data and the corresponding semantic feature vector of the urban safety historical data. The expression is: , where is the semantic feature vector of the newly generated urban safety historical data and the th semantic feature vector of the urban safety historical data The semantic similarity, is and The Manhattan distance, is and all The maximum Manhattan distance, is The word frequency length of the corresponding semantic data, is The word frequency length of the corresponding semantic data, is all The maximum length of the word frequency of the corresponding semantic data; Calculate the numerical similarity between the newly generated urban safety historical data and the corresponding numerical feature vector of the urban safety historical data. The expression is: , Among them is the numerical feature vector of the newly generated urban safety historical data and the th numerical feature vector of the urban safety historical data has a numerical similarity , are respectively the th numerical element value of the newly generated numerical feature vector of the urban safety historical data and the mean value of all numerical elements , are respectively the th numerical element value of the numerical feature vector of the urban safety historical data and the mean value of all numerical elements is the dimension of the feature vector Thus, the similarity between the newly generated urban safety historical data and the th group of urban safety historical data is calculated as , is the similarity weight In actual evaluation, taking the newly generated urban safety historical data of a certain city as an example, 3, 2, 4, 1, and 5 groups of urban safety historical data to be replaced are respectively determined in the categories of natural disasters, accident disasters, public safety, public health, and infrastructure. The whole group of urban safety historical data to be replaced is extracted according to the time stamp, and there are 6 groups of duplicate data among them. The overall similarities between the newly generated urban safety historical data and the remaining 9 groups of urban safety historical data to be replaced are calculated: 0.72, 0.85, 0.78, 0.68, 0.81, 0.76, 0.65, 0.89, 0.74. Five groups of urban safety historical data update groups are determined by the overall similarity threshold of 0.75, and the corresponding time series weights are 0.3, 0.4, 0.6, 0.6, 0.9. The time series overall similarities are calculated as 0.255, 0.312, 0.486, 0.456, 0.801. The urban safety historical data with a time series overall similarity of 0.255 is replaced with the newly generated urban safety historical data;

[0018] In this embodiment, the method for determining the evaluation reference value includes: Obtain the urban safety data to be evaluated, classify the urban safety data to be evaluated by using a decision tree, screen the classification results according to the risk thresholds of different types of urban safety data to obtain risk data, and extract the historical deviations of different types of data to obtain a historical deviation vector; the historical deviation is determined by the historical maximum value, historical minimum value, and historical mean value; Construct a historical data matrix based on the urban database; the column data of the historical data matrix is used to represent different statistical forms of the same type of data; the row data of the historical data matrix is used to represent different types of data in the same statistical form; the different statistical forms include median, mode, quartile, mean, maximum value, minimum value, range, skewness, and kurtosis; Sort the risk data by category to obtain a risk vector, compare the risk vector with the historical data matrix to obtain a historical difference matrix, determine a risk threshold vector based on the risk thresholds of different types of urban safety data, divide each element in each row of the historical difference matrix by the corresponding element in the risk threshold vector to obtain a weighted difference coefficient matrix, and determine the evaluation benchmark value for each category according to the weighted difference coefficient matrix and the historical deviation vector; In actual evaluation, the risk thresholds of urban safety data are determined by statistical methods as follows: 1. Natural disasters (category 1 typhoon, typhoon affecting 10% of the urban area, daily rainfall of 50 mm, riverbed height increased by 1.5 m, number of waterlogging areas 3, waterlogging depth 0.1 m), 2. Accident disasters (no accident disaster events), 3. Public health (no special events), 4. Infrastructure (no old house collapses, 2 urban elevator malfunctions, 5 power operation malfunctions, no gas leakage and explosion events, no bridge operation malfunction events, 1 water supply and drainage pipe network failure, normal traffic congestion index, normal number of emergency service calls, 20% increase in mobile network traffic, 1 public facility damage), 5. Public safety (3 public transportation malfunctions, 3 theft events); Take the newly generated urban safety historical data as the input urban safety data, screen to obtain risk data, and obtain the evaluation benchmark values of 3.5, 3.5, 3, 3, 3 for each category of data according to the classification weights of urban safety historical data and the historical data matrix.

[0019] In this embodiment, the method for constructing the risk assessment layer includes: Calculate the average value of the data correlation between risk data of the same category and risk data of other categories to obtain a risk weight; Calculate the deviation value between risk data of different categories and the corresponding risk thresholds to obtain a risk deviation; Calculate the prior probability of urban safety data according to the historical data matrix and associated events, and use Bayes' theorem to calculate the corresponding risk probability according to risk data of different categories and the prior probability. Calculate the prior probability of urban safety data according to the historical data matrix and associated events, and use Bayes' theorem to calculate the corresponding risk probability according to the risk data and the prior probability Determine the risk valuation of different types of urban safety data according to the risk weight, risk deviation, and risk probability. The expression of the risk valuation function is: , where For the risk valuation of the category of urban safety data, for the bias term of the risk valuation of the category of urban safety data risk valuation, used to adjust the valuation benchmark, for the risk weight of the category of risk data, for the th data index and the risk deviation from the corresponding risk threshold in the category of risk data, for the number of data indexes in the category of risk data, for the risk probability of the 、 for the mean and standard deviation of the prior probability of the category of urban safety historical data, for the regularization coefficient, is a non - zero constant; In actual evaluation, calculate the mean of data correlation to obtain the risk weights of various types of data (natural disaster type, accident disaster type, public health type, infrastructure type, public safety type): 0.4, 0.2, 0.1, 0.2, 0.1; Calculate the mean and standard deviation of the prior probability of urban safety data as 0.22, 0.18, 0.20, 0.24, 0.19 and 0.03, 0.04, 0.03, 0.02, 0.03 respectively according to the historical data matrix and associated events; Calculate the risk probability according to Bayes' theorem: 0.4, 0.35, 0.33, 0.33, 0.3; The bias terms of risk valuation are taken as 5, 4, 3, 3, 3 respectively, and calculate the risk valuation according to the risk valuation function as: 5.898, 4.2, 3, 3.5, 3.2; At this time, conduct risk early warnings for natural disasters and infrastructure - type risks according to the risk valuation deviation.

[0020] In this embodiment, the method for optimizing the model includes: Determine the objective function for optimizing the early - warning model according to the risk valuation deviation, and the expression is: , where is the objective function for optimizing the early - warning model, is the evaluation benchmark value of the category of urban safety data, is the number of categories of urban safety data, is the number of indexes of the category of urban safety data, is the regularization parameter, is the hyper - parameter of the weight of the interpretive term, is the parameters, being the number of model parameters; The existing parameter information is updated using Gaussian process regression, and the expression is: , where is the mean value of the function at , is the input point corresponding to the risk valuation function, is the non - linear mapping of the input data in the high - dimensional feature space, is the covariance matrix of the training data in the mapped feature space, calculated by the adaptive kernel function, is the non - linear mapping of the training data in the high - dimensional feature space, is the variance of the data Gaussian noise, is the identity matrix, is the target vector of the training data, is the covariance of the function at , is the adaptive kernel function, is the kernel function parameter for adaptive adjustment; The most promising set of hyperparameters is selected for the next evaluation according to the acquisition function, and the acquisition function expression is: , where is the hyperparameter selection benchmark, is the target function threshold, is the proposed hyperparameter, is the actual value of the target function when using the hyperparameter , is the actual value of the target function being the probability of the surrogate model, is the target function value less than the threshold when the hyperparameter appears, is the target function value less than the threshold probability, is the target function value greater than the threshold when the hyperparameter appears; Update the parameters and evaluate the hyperparameters. Repeat the above operations until the hyperparameters are no longer updated when the objective function value of the early - warning model optimization is minimized, and output the urban safety risk assessment benchmark early - warning model; Input the urban safety data to be evaluated into the optimized urban safety risk assessment benchmark warning model to obtain the risk valuations and corresponding assessment benchmark values of different categories of urban safety data. Conduct risk warnings based on the risk valuation deviations, and use the corresponding risk data, risk thresholds, assessment benchmark values, and risk valuations as the historical assessment set, and store the historical assessment set in partitions.

[0021] In a second aspect, an early warning model system for an urban safety risk assessment benchmark includes: Urban database module: used to store urban safety historical data, calculate the similarity of the urban safety historical data to update the urban safety historical data, determine the historical data matrix based on the urban safety historical data, determine the risk threshold, and store the historical assessment set; Risk identification module: used to obtain urban safety data and screen the urban safety data according to the risk threshold to determine the risk data; Assessment benchmark value module: used to obtain the historical data matrix according to the urban database and determine the assessment benchmark value based on the historical matrix and the risk data; Risk assessment module: used to calculate the risk weight, risk deviation, and risk probability, and determine the risk valuations of different categories of urban safety data according to the risk weight, the risk deviation, and the risk probability Model optimization module: used to optimize the urban safety risk assessment benchmark warning model according to the risk valuation deviation; Risk management module: used to view and manage the urban safety historical data and the historical assessment set, and conduct risk warnings according to the risk valuation and the assessment benchmark value.

[0022] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for constructing an early warning model for an urban safety risk assessment benchmark, characterized in that: The following steps are involved: S1. Build a city database: Update historical city safety data and determine risk thresholds; S2. Building a risk identification layer: determining risk data according to the risk threshold; S3, determining the assessment benchmark value: obtaining a historical data matrix according to the city database, and determining the assessment benchmark value according to the historical data matrix and the risk data; S4. Constructing a risk assessment layer: determining a risk valuation based on the risk data and the risk threshold; S5. Optimization model: Optimize the urban safety risk assessment benchmark warning model based on risk valuation deviation; S6. Constructing a risk warning layer: conducting risk warning according to the risk valuation and the assessment benchmark value.

2. The method for constructing an early warning model for a city safety risk assessment benchmark according to claim 1, characterized in that: The method for constructing the city database comprises: Obtain the city safety historical data, classify the data using clustering method, partition and store the city safety historical data according to the classification results, and store the city safety historical data in storage partitions in segments according to the collection time; Set up a preliminary screening window for urban safety data, determine the parameters of the preliminary screening window, and perform sliding updates on the urban safety historical data in the urban database in the preliminary screening window for urban safety data according to the time series; the preliminary screening window parameters include update frequency and window length; Determine the update weights of various types of data, and divide the various types of data into first-class data and second-class data according to the data update weights; Acquire newly generated city safety historical data, and determine the data preliminary screening result according to the data deviation between the newly generated city safety historical data and the mean value of the city safety historical data in the city safety data preliminary screening window; the data preliminary screening result includes further determining the data update status and whether to update the historical data; The steps to further determine the data update status include: Calculate the category similarity between the newly generated urban safety history data in different partitions and the urban safety history data in the category, determine the urban safety history data to be replaced in the partition according to the category similarity threshold, and extract the corresponding urban safety history data group according to the timestamp; Calculate the overall similarity between the newly generated urban safety history data and the urban safety history data in all categories, determine the urban safety history data update group according to the overall similarity threshold, and when the number of update groups is less than 3, store the newly generated urban safety history data in the city database according to the rules and do not replace the urban safety history data. Otherwise, add the time series weight to calculate the overall time series similarity, and use the newly generated urban safety history data to replace the urban safety history data with the lowest overall time series similarity; Based on the updated city database, statistical methods are used to determine the risk threshold of urban safety data.

3. The method for constructing an early warning model for an urban safety risk assessment benchmark according to claim 2 is characterized in that: The method for calculating the similarity includes: Distinguish between numerical data and semantic data, and set semantic evaluation sets , is the evaluation set element corresponding to the semantic evaluation of the semantic data, To evaluate the degree, the semantic function Numerical transformation is performed to obtain semantic numerical data, and semantic feature vectors are determined based on the semantic numerical data. The semantic similarity between the newly generated urban safety history data and the semantic feature vectors corresponding to the urban safety history data is calculated. The expression is: ， in is the semantic feature vector of the newly generated urban safety historical data With Semantic feature vector of city safety history data The semantic similarity of for and The Manhattan distance of for With all The maximum Manhattan distance of for The word frequency length corresponding to the semantic data, for The word frequency length corresponding to the semantic data, For all The maximum length of the word frequency corresponding to the semantic data in; Calculate the numerical similarity between the newly generated urban safety historical data and the corresponding numerical feature vector of the urban safety historical data. The expression is: ， in is the numerical feature vector of the newly generated urban safety historical data With Numerical feature vector of city safety history data The numerical similarity of , are the numerical feature vectors of the newly generated urban safety historical data No. The value of the numeric elements and the mean of all numeric elements, , are the numerical feature vectors of urban safety historical data No. The value of the numeric elements and the mean of all numeric elements, is the dimension of the feature vector; The newly generated urban safety historical data is calculated and The similarity of the safety history data of the group of cities is , is the similarity weight.

4. The method for constructing an early warning model for an urban safety risk assessment benchmark according to claim 1, characterized in that: The method for determining the evaluation benchmark value includes: Obtain the city safety data to be evaluated, classify the city safety data to be evaluated using a decision tree, screen the classification results according to the risk thresholds of different types of city safety data to obtain risk data, and extract the historical deviations of different types of data to obtain a historical deviation vector; the historical deviation is determined by a historical maximum value, a historical minimum value, and a historical mean value; A historical data matrix is ​​constructed according to the city database; the column data of the historical data matrix is ​​used to represent different statistical forms of the same type of data; the row data of the historical data matrix is ​​used to represent different types of data in the same statistical form; Sort the risk data by category to obtain a risk vector, compare the risk vector with the historical data matrix to obtain a historical difference matrix, determine the risk threshold vector based on the risk thresholds of different types of urban safety data, divide each row of the historical difference matrix by the corresponding element of the risk threshold vector to obtain a weighted difference coefficient matrix, and determine the assessment benchmark value for each category based on the weighted difference coefficient matrix and the historical deviation vector.

5. The method for constructing an early warning model for urban safety risk assessment benchmark according to claim 1, characterized in that: The method for constructing the risk assessment layer comprises: Calculate the mean value of the correlation between the same category of risk data and other categories of risk data to obtain the risk weight; Calculate the deviation between different categories of risk data and corresponding risk thresholds to obtain risk deviation; Calculate the prior probability of urban safety data based on the historical data matrix and related events, and use Bayesian theorem to calculate the corresponding risk probability based on risk data and prior probability of different categories. Calculate the prior probability of urban safety data based on the historical data matrix and related events, and use Bayesian theorem to calculate the corresponding risk probability based on risk data and prior probability; The risk valuation of different types of urban safety data is determined based on risk weight, risk deviation and risk probability. The risk valuation function expression is: ， in For the Risk valuation of city safety data, For the The bias term of the risk valuation of the city safety data is used to adjust the valuation benchmark. For the Risk weights for risk-based data, For the The risk data The risk deviation between each data indicator and the corresponding risk threshold, For the The number of data indicators of risk-related data, For the The risk probability of class risk data, , For the The mean and baseline difference of the prior probability of the historical safety data of similar cities, is the regularization coefficient, is a non-zero constant.

6. The method for constructing an early warning model for urban safety risk assessment benchmark according to claim 1 is characterized in that: The method for optimizing the model comprises: The objective function of early warning model optimization is determined according to the risk valuation deviation, and the expression is: ， in is the objective function optimized by the early warning model, For the The assessment benchmark value of the safety data of this type of city, is the number of categories of urban safety data, For the The number of indicators of urban safety data, is the regularization parameter, is the hyperparameter of the explanatory term weight, For the model parameters, is the number of model parameters; Gaussian process regression is used to update the existing parameter information, and the expression is: ， in For the function The mean value of is the input point corresponding to the risk valuation function, is the nonlinear mapping of input data in high-dimensional feature space, is the covariance matrix of the training data in the feature space after mapping, which is calculated by the adaptive kernel function. is the nonlinear mapping of training data in high-dimensional feature space, is the data Gaussian noise variance, is the identity matrix, is the training data target vector, For the function The covariance of is the adaptive kernel function, is the kernel function parameter, used for adaptive adjustment; The most promising set of hyperparameters is selected for the next evaluation according to the acquisition function, which is expressed as: ， in Selecting benchmarks for hyperparameters, is the objective function threshold, are the proposed hyperparameters, To use hyperparameters The actual value of the objective function is The actual value of the objective function is The probability of the proxy model when is the objective function value Less than threshold Hyperparameters The probability of occurrence, is the objective function value Less than threshold The probability of is the objective function value Greater than threshold Hyperparameters Probability of occurrence; Update parameters and evaluate hyperparameters, repeat the above operations until the value of the objective function optimized by the early warning model is minimized, stop updating hyperparameters, and output the urban safety risk assessment benchmark early warning model; The urban safety data to be assessed is input into the optimized urban safety risk assessment benchmark early warning model to obtain risk valuations and corresponding assessment benchmark values ​​for different categories of urban safety data. Risk warnings are issued based on risk valuation deviations, and the corresponding risk data, risk thresholds, assessment benchmark values ​​and risk valuations are used as historical assessment sets, which are stored in partitions.

7. An early warning system for urban safety risk assessment benchmark, used to execute the method according to any one of claims 1 to 6, characterized in that: include: City database module: used for storing city safety historical data, calculating the similarity of the city safety historical data and updating the city safety historical data, determining the historical data matrix according to the city safety historical data, determining the risk threshold, and storing the historical assessment set; Risk identification module: used for obtaining city safety data, screening the city safety data according to the risk threshold to determine risk data; An assessment benchmark value module: used to obtain a historical data matrix according to the city database, and determine the assessment benchmark value according to the historical data matrix and the risk data; Risk assessment module: used to calculate risk weight, risk deviation and risk probability, and determine the risk valuation of different categories of urban safety data based on the risk weight, risk deviation and risk probability Model optimization module: used to optimize the urban safety risk assessment benchmark warning model according to the risk valuation deviation; Risk management module: used to view and manage the city safety historical data and the historical assessment set, and to issue risk warnings based on the risk valuation and the assessment benchmark value.

Images